This investigation sought to provide the first prospective examination of the association between the presence of a current cannabis use disorder (CUD) diagnosis and changes in posttraumatic stress disorder (PTSD) symptoms over time after discontinuation, after accounting for the theoretically relevant effects of age, trauma severity, psychological distress, and co-occurring substance use disorders. The study was longitudinal and assessed clinical patients at two time points: residential PTSD treatment intake and discharge. The sample consisted of 260 male combat-exposed military veteran patients (Mage = 52.57 years, SD = 5.47) admitted to a Veterans Affairs residential rehabilitation program for PTSD between 2000 and 2008. Diagnoses were obtained using the Structured Clinical Interview for DSM–IV, and PTSD outcomes were determined by the PTSD Checklist—Military Version. Results indicate that the presence of a CUD diagnosis was significantly predictive of lower levels of change (between treatment intake and discharge) in PTSD symptom severity as well as PTSD avoidance–numbing and hyperarousal symptom cluster severity (all ps < .05). This study presents the first evidence of a prospective relation between problematic cannabis use and PTSD symptoms, indicating that individuals with a CUD are likely to experience lower levels of change in PTSD symptoms over time, within the context of discontinuation owing to residential PTSD treatment.